I volunteered at SolarSPELL (offline library program), an educational initiative undertaken at Arizona State University during the summer of 2020. SolarSPELL aims to bring the experience of using the internet and digital educational content to remote areas of the world. They distribute the educational content using a solar-powered micro-controller that hosts a local wi-fi-based service. This service can be accessed by users using any device with a web-browser from any location where it is deployed.
The team at SolarSPELL was in the process of deploying a dedicated library for a nursing-midwifery school in South Sudan. Unlike their previous libraries, which were mostly designed for young learners, the nursing library was focused on adult learners. I was designing a new experience for this new project. The initial research findings made me focus on content discovery. The library for young learners was designed to be used in a controlled environment with a structured use. It had limited search capabilities (query doesn’t return results from sub-folder in a file location), which was mainly due to the technical limitations of the micro-controller used.
Search option is not the answer to all content-discovery user goals
I found through a survey that adult learners would be using this library for three different purposes:
(i) getting additional info on a specific topic
(ii) exploring new content
(iii) learning a new subject.
We had to explore options beyond the catalog-like presentation and search capabilities for satisfying user needs. Unlike young learners, adult learners will need features that support the free exploration of the available content. While the existing limitation of search can be fixed, catalog-listing and search require the user to have previous knowledge of what they are seeking. This means, for users who want to learn something specific but don't know about it yet, finding the content would be difficult or impossible in a few instances.
Based on user research and doing secondary research on existing content-discovery options, we found features like tags and index. These will serve the opportunities we have found. To an untrained eye, this might feel like serving a similar function but they have clear distinctions which help users with different content discovery goals.
Index/glossary
While testing the existing library for usability issues, I found that similar content was available under different topics and was not connected by any means. For example, unique content on “solar” was available under “Environment>Energy” & “Science>Renewable Energy.” Unless the user was aware of the entire catalog, they would not find this content that might be of interest to them. The library had a content hierarchy with too many subsections (some of the content required users to click through 6 sections to reach the content)and the unavailability of a global search function also added to the content accessibility issue. Easy discovery of similar content is an important aspect to have in the offline library as it has limited resources; unlike the internet, all the content with a similar topic should be easily accessible.
The index was a perfect feature for the medical library. I stumbled on this idea while talking to a pseudo-user (a medical student from India). Medical books usually have an index with the keywords and their occurrence listed at the end of the book. Since the library is designed for nursing students, this would meet their expectations and would not require additional info for using the feature as they would be familiar with the concept.
This feature coupled with a new type of search would also solve the issue of homonyms (similar-sounding words) in medical terms. English was a second language for the majority of users so the ability to search for a specific topic with partial spelling, common suffix, or prefix would also be helpful. By making index-search behave like the find feature on a browser, users can search for terms using partial words.
Tags and content suggestion
All the content (PDFs and media files) in the library is curated by a content specialist. In the process of curation, every file gets a title, description, and tags added to it which helps to implement the index feature. These tags can also help the user to discover content in more useful ways. We are accustomed to exploring content using hashtags. SolarSPELL also aims to provide the experience of using the internet to people who are yet to experience it. By allowing users to explore content using tags, they will pick up an internet skill that might help them understand a lot of new internet-based apps and services when they use them in the future.
On the internet, we get advanced content suggestions based on multiple factors. We may not be able to provide such an advanced feature for an offline experience due to hardware limitations but since we will have tags for the uploaded content, we can display a few of them to mimic the internet features. This also has a down-side; we might end up displaying the same content again and again, which might feel frustrating to experienced users who may have used the library for a long time.
Implementation
Based on the research, I sketched and created prototypes to test with real users. I have provided the examples below from the first version of the prototype that reflects content from children’s library, as the medical library was still being curated. So forgive me and kindly ignore the content (it’s a UX sin not to have the actual content in the prototype, but it's better than Lorem ipsum).
Content Suggestion
Tags
Index feature
Index-search
Index-search results
Global-site Search
Final thoughts
This project helped me realize that technical limitations should not stop designers from crafting new experiences for the users. While the quality of content is important for educational resources, it is equally important for us to ensure the ease of access to the available content, especially when the resource available is limited. This increases the value we can deliver to the end-users by providing multiple pathways to discover the content.
P.S. The solution showcased here is the first version of the prototype, this can still be improved further by how it is presented to the user, example: instead of the headings like “similar content suggestion” & “related titles”, we can use “You may also like” which sounds more natural to the user. I’m pointing this out just as a reminder that UX can be improved independently from the technology used.
